Research on Enterprise Production Decision Making Based on Genetic Algorithm and Monte Carlo Simulation

This paper launched an in-depth study on quality inspection and cost control in the production process of enterprises. Firstly, the right-tailed one-sided test is used to determine the sample sizes at 95% and 90% confidence level, which are 138 and 98 respectively, by defining the original and alternative hypotheses and using the confidence level and critical value calculation formulas in statistics. Next, the decision variables and parameters are defined to optimize the purchasing and assembling strategies of spare parts and finished products by Genetic Algorithm (GA). Assuming that the purchasing quantity of spare parts 1 and 2 is 100 and the assembly quantity of finished products is 80, it is finally found that no testing can effectively save costs, and the minimum total cost is $617.56. In addition, Monte Carlo simulation is chosen to cope with the complex and multi-stage problem with uncertainty. By randomly generating the substandard status of each spare part and finished product, adjusting according to the predetermined strategy, calculating each cost, and summarizing the results after simulating several times. The comprehensive analysis generates 499 decision scenarios, which are visualized by scatter plots and histograms to help enterprises trade-off between defective rate and production cost and choose the optimal strategy.